Activation of IGF1R/p110β/AKT/mTOR confers resistance to α-specific PI3K inhibition

Human cell lines T47D and MCF7 were obtained from the American Type Culture Collection (Manassas, VA, USA), were authenticated by single nucleotide polymorphism fingerprinting, and were generally used within 20 passages. The cell lines were maintained in RPMI medium supplemented with 10 % fetal bovine serum, 10 μg/ml human insulin solution, 100 IU/ml penicillin, and 100 μg/ml streptomycin. All lines were maintained at 37 °C with 5 % CO₂. To develop resistant models, parental cell lines were chronically treated with IC90: (90 % inhibitory concentrations) of BYL719 (2 μM for T47D and 5 μM for MCF7) over a period of 5–6 months until resistance occurred. Fresh media were provided every 2 days.

Antibodies used in this study were anti-pAKT Ser473, anti-AKT, anti-pS6 Ser235/236, anti-S6, anti-pPRAS40 Thr246, anti-pERK1/2 Thr202/Tyr204, anti-ERK1/2, anti-p110α, anti-p110β, anti-PTEN, anti-IRS1 (Insulin receptor substrate), anti-poly (ADP-ribose) polymerase (anti-PARP), and anti-cleaved PARP from Cell Signaling Technology (Danvers, MA, USA), anti-phosphotyrosine 4G10 and anti-p85 from Millipore (Billerica MA, USA) anti-pIGF1R/IR Tyr1162/1163 from Biosource (ThermoFisher Scientific, Waltham MA, USA) and anti-IRS2 from Novus Biological (Littleton CO, USA). BYL719, AEW541, RAD001, and MEK162 were provided by Novartis Pharma AG (Basil, Switzerland). MK2206, AZD6482, GSK2334470, CAL101, and BMS354825 were purchased from Selleckchem (Houston TX, USA). All compounds used in vitro were dissolved in dimethyl sulfoxide (DMSO).

Cells were harvested in lysis buffer (200 mM ammonium bicarbonate, pH 7.5, 8 M urea) supplemented with the PhosStop phosphatase inhibitor cocktail Roche Diagnostics (Rotkreuz, Switzerland), reduced and alkylated, and then digested with Trypsin/LysC Promega (Madison WI, USA) after dilution to 2 M urea. The peptides were acidified to 1 % TFA (trifluoroacetic acid) and desalted on SepPak C18 cartridges. Lyophilized peptides were dissolved in immunoprecipitation buffer (50 mM ammonium bicarbonate, pH 7.4, 150 mM NaCl, 1 % (w/v) octyl-β-d-glucopyranoside, and Roche protease and phosphatase inhibitors (Complete and PhosStop, Roche Diagnostics, Rotkreuz, Switzerland) and incubated with 200 μl of anti-phosphotyrosine antibodies (PY99; SantaCruz, Dallas TX, USA) for 16 hours at 4 °C. After elution of the beads with 0.1 % TFA, peptides were desalted on Poros R3 and further purified on TiO₂ microcolumns.

The purified phosphopeptides were resuspended in 10 % formic acid and injected onto a 15 cm × 75 μm ProteoPep 2 PicoFrit column (New Objectives, Woburn MA, USA) connected to an LTQ-OrbiTrap Elite mass spectrometer Thermo (ThermoFisher Scientific, Waltham MA, USA). Buffer A consisted of H₂O with 0.1 % formic acid and Buffer B of 100 % acetonitrile with 0.1 % formic acid. Peptides were separated using a two-step gradient, from 0 % B to 20 % B in 90 minutes and from 20 % B to 50 % B in 50 minutes. Data acquisition used a “Top 15 method”, where every full mass spectrometry (MS) scan was followed by 15 data-dependent scans on the 15 most intense ions from the parent scan. Full scans were performed in the OrbiTrap at 120,000 resolution with target values of 1E6 ions and 500 milliseconds injection time, while MS/MS scans were performed in the ion trap with 1E4 ions and 200 milliseconds injection time. Database searches were carried out with Mascot Server (Matrix Science, London UK) using the human Uniprot database (release 20130429, www.​uniprot.​org). Mass tolerances were set at 10 ppm for the precursor and at 0.8 Da for the fragment ions. In the case of ambiguous assignments, spectra were manually interpreted for confirmation of identity and localization of the phosphorylation site using Scaffold (version 4.3; Proteome Software, Portland OR, USA). Label-free quantification was performed on duplicate liquid chromatography (LC)–MS runs for each sample using Progenesis LC-MS (Nonlinear Dynamics Software, Newcastle-upon-Tyne, UK). Normalized peptide intensities were added together for each unique phosphorylated peptide with Mascot scores exceeding 20, and used to calculate the log₂ ratios between samples for each unique phosphopeptide.

Total proteins were extracted with 50 mM Tris–HCl, pH 7.6, 150 mM NaCl, 1 % NP40, and Roche phosphatase and protease inhibitor cocktail. Immunoprecipitation was performed overnight at 4 °C on total protein lysates with the indicated antibodies according to the supplier’s recommendations. Protein G sepharose beads were then added for 1 hour at 4 °C. Proteins were eluted with 50 μl loading buffer (45 μl LDS sample buffer + 5 μl Reducing Agent; Novex, ThermoFisher Scientific, Waltham MA, USA,) and then boiled at 95 °C for 5 minutes. Immunoprecipitates or 50 μg of proteins for a whole cell lysate were loaded onto 4–12 % SDS-PAGE gels using the NuPAGE system from Invitrogen (ThermoFisher Scientific, Waltham MA, USA) (30 mA for 10 minutes and 50 mA for 1 hour 15 minutes) and then transferred onto Invitrolon PVDF membranes in the Biorad Blotter system (Hercules CA, USA), using blotting buffer containing 25 mM Tris-Base, 192 mM glycine, and 5 % methanol (100 V for 30 minutes). Detection was by chemiluminescence using the Western Bright ECL detection kit Advansta (Menlo Park CA, USA). Immunoblots are representative of a minimum of three independent experiments. An equal amount of protein was loaded onto each gel and the immunoblots shown in the same figure were developed simultaneously with the same exposure time.

Standard cell growth was performed in 2 % fetal calf serum medium with appropriate concentrations of the indicated inhibitors using 24-well plates (50,000 cells/well) and measured by sulforhodamide B staining according to the manufacturer’s instructions (Sigma Aldrich, St. Louis MO, USA). Dose-response experiments were evaluated at day 3 and time course experiments at days 3, 6, and 9. In the time course experiments, cells were plated and treatments with DMSO vehicle (VHC) or the indicated inhibitors started 6 hours later. VHC-treated resistant cell lines do not have BYL719 in the media all through the experiment.

A total of 70,000 cells/well were seeded in six-well plates. After overnight incubation, media were aspirated and replaced with media with the indicated drugs. After 72 hours, the media were collected and cells were harvested, washed twice with cold phosphate-buffered saline (PBS), and resuspended in Annexin binding buffer. Cells were stained with propidium iodide (PI) and Annexin V according to the manufacturer’s protocol (BD Biosciences, San Jose, CA, USA).

Female Balb-c nude mice were used in compliance with the Swiss laws on animal welfare and the animal protocols were approved by the Swiss Cantonal veterinary Office of Basel. MCF7 cells (5 × 10⁶ cells) were suspended in a 50 μl mixture of Matrigel (BD Biosciences, San Jose CA, USA) and PBS (1/1), and were injected subcutaneously into the right flank of female Balb-c nude mice 6–8 weeks old. Mice were implanted with estrogen pellets on the day of cell injection. Tumor-bearing mice were randomized into six groups of five or six mice based on tumor volume prior to initiation of treatment, which started when the average tumor volume was between 150 and 200 mm³. BYL719, AEW541, and GSK2636771 were given orally daily. BYL719 and GSK2636771 were dissolved in CMC/Tween and AEW541 in NMP/PEG300 (1/9). VHC-treated mice received a combination of CMC/Tween and NMP/PEG300. Tumors were measured every 5 days and tumor volumes calculated by the formula:

For dose-response experiments, one-way analysis of variance analysis was performed using GraphPad Prism (Graphpad, La Jolla CA, USA) 6.0 software to determine GI50 (50 % growth inhibition) growth inhibition values. For all analyses, reported values represent the means ± standard error of the mean (SEM) of at least three independent experiments. Data were tested with Student’s t test and P <0.05 was considered statistically significant.